Quantum software development lacks the systematic experiment tracking found in mature fields, leaving results vulnerable to irreproducibility. This work identifies dimensions, circuit parameters, measurement results, and noise characteristics, absent from classical machine learning pipelines, demanding bespoke tooling for quantum programs. By demonstrating this with quantum reservoir computing, a pathway towards reliable quantum software emerges.
Iet Research Develops Experiment Tracking Protocol for Quantum Software Development
Quantum software development lacks the systematic experiment tracking found in mature fields, leaving results vulnerable to irreproducibility. This work identifies dimensions, circuit parameters, measurement results, and noise characteristics, absent from classical machine learning pipelines, demanding bespoke tooling for quantum programs. By demonstrating this with quantum reservoir computing, a pathway towards reliable quantum software emerges.